Skip to content

Contributing to analyses

This section contains information about how to create, contribute to, and/or run an analysis module using the Git/GitHub framework.

All instructions in this section assume you have already:

  1. Been onboarded to the project as a new contributor
  2. Taken all set up steps to work on the project
  3. Discussed your proposed analysis on GitHub
  4. Filed an issue that tracks steps you'll take to perform your analysis

Once you have an issue to work off of, you're ready to go!

Project Overview

The image below provides a high-level overview of the Git/GitHub aspects of the project workflow.

Overview of the project flow

You will contribute to the project by creating and developing analysis modules.

Before you begin

Before you create your module, please review the documentation in the Analysis Module and Managing software dependencies sections.

This information will help orient you to how to write OpenScPCA analysis modules and ensure they meet the project's reproducibility requirements.

The below steps outline the process for getting started with a new analysis module.

  1. Create a feature branch in your fork to begin your analysis (but see the helpful tips below!)
  2. Create a new analysis module (unless you are working on an existing module)
  3. Perform your analysis using Git:
    • Commit small units of work, pushing your commits to your fork on GitHub along the way
    • Be sure to document your analysis module as you build it up
  4. When you've completed enough work for a pull request (PR), you'll file a PR to the upstream repository, AlexsLemonade/OpenScPCA-analysis
  5. The Data Lab will then review your PR
  6. Once we approve your PR, we will merge your feature branch into the upstream AlexsLemonade/OpenScPCA-analysis repository
  7. Finally, you'll sync your fork with the upstream repository
    • Then, rinse and repeat - make your next feature branch and begin again!

Helpful tips

  • Before you dive into an analysis, take some time to plan your work.

    • Focused units of work will make the pull request process go smoothly and quickly!
    • You should expect that for each analysis you work on, you will go through several iterations of feature branches and PRs. You'll save time and energy in the long-run by thinking a little bit about what each branch, and therefore each PR, will contain.

  • Before you file a pull request, we strongly recommend syncing your branch with the upstream repository.

    • This will reduce odds of having a merge conflict, which can be tricky to deal with (but we're here to help!)
    • You can also periodically sync up while you're working on your analysis to even further reduce the chances of having a merge conflict

  • Remember, we're here to help!

    • Refer to our Getting Help section at any time to learn about different ways to get help from the Data Lab as you work on your analysis.